A new study from University of Washington (UW) has shown that machine learning can be used to improve forecasts for lightning.
Lightning is a destructive force that has the potential to cause extensive damage to infrastructure, buildings and even create huge fires such as the massive California Lightning Complex fires. Having the ability to prepare for potential lightning forecasts could lead to better readiness for wildfires, improve warning times and create longer climate models.
“The best subjects for machine learning are things that we don’t fully understand,” said Daehyun Kim, an associate professor of atmospheric sciences at UW. “And what is something in the atmospheric sciences field that remains poorly understood? Lightning. To our knowledge, our work is the first to demonstrate that machine learning algorithms can work for lightning.”
How it works
The technology uses weather forecasts combined with a machine learning equation based on previous lightning events. The method allows researchers to forecast lightning over a region two days earlier than existing methods.
“This demonstrates that forecasts of severe weather systems, such as thunderstorms, can be improved by using methods based on machine learning,” said Wei-Yi Cheng, a UW doctorate student in atmospheric sciences who worked on the project. “It encourages the exploration of machine learning methods for other types of severe weather forecasts, such as tornadoes or hailstorms.”
Researchers trained the system with lightning data from 2010 to 2016 and let the system discover relationships between variables and lightning bolts. It was then tested on weather from 2017 to 2019, comparing the artificial intelligence (AI)-supported technique and an existing physics-based method.
The UW method was able to forecast lightning with about the same skill about two days earlier than the older method. In the southeastern U.S., where lightning is frequent, the AI was found to be more accurate than other methods but in locations where lightning is sparse it was not as accurate.
The next steps include using more data sources, more weather variables and sophisticated technologies to improve predictions or particular situations like dry lightning or lightning without rainfall.
Eventually, UW said the method could be used for longer-range projections, which is important because lightning affects air chemistry which in turn would lead to better climate models by predicting lightning.
“In atmospheric sciences, as in other sciences, some people are still skeptical about the use of machine learning algorithms — because as scientists, we don’t trust something we don’t understand,” Kim said. “I was one of the skeptics, but after seeing the results in this and other studies, I am convinced.”
